Interactive Machine Learning

Satori

Understanding through visualization

Explore complex machine learning concepts through stunning interactive experiences. From Gaussian Splatting to Neural Implicit Surfaces — build deep intuition by seeing and interacting with the math.

Start Exploring

Visualization Topics

Choose your journey

Each topic builds from foundational concepts to cutting-edge techniques. Interact with every visualization to develop true understanding.

Fundamentals

2D Gaussian Splatting

Start with the basics. Understand how 2D Gaussians represent visual information — manipulate position, covariance, color, and opacity in real time to build intuition for the core primitive.

Start exploring

Scene Reconstruction

3D Gaussian Splatting

See how millions of 3D Gaussians compose into photorealistic scenes. Explore the rasterization pipeline, spherical harmonics for view-dependent color, and the differentiable rendering process.

Start exploring

Neural Implicit Surfaces

NeUS & Supernormals

Explore how NeUS learns signed distance functions from images. Visualize how surface normals are extracted and refined, and understand the supernormal estimation process.

Start exploring

Digital Humans

Person-Specific Gaussians

Discover how Gaussian Splatting is adapted for human avatars. Explore pose-dependent deformations, expression blending, and real-time rendering of photorealistic digital humans.

Start exploring